DocumentCode :
2636963
Title :
Precipitation State Forecasting Based on Unascertained CMeans and Markov Chain Model with Gray Relevancy Weights
Author :
Li-Hua Ma ; Hui-Zhe Yan
Author_Institution :
Sch. of Econ. & Manage., Hebei Univ. of Eng., Handan
fYear :
2008
fDate :
18-20 June 2008
Firstpage :
321
Lastpage :
321
Abstract :
At present, in the field of hydrology and meteorological science, precipitation state forecasting is an extremely important problem. In this paper, the problem of precipitation state forecasting was studied, and a new forecasting method based unascertained c-means and Markov chain model with gray relevancy weights was presented. The method included the unascertained characteristic of precipitation state comprehensively, thus its forecasting outcomes are more scientific. Firstly, the unascertained C-means method is applied to divide time series of precipitation state, and the unascertained classification standard of precipitation state is established based on the fact that there are a lot of unascertained characteristics in the precipitation. Secondly, a forecasting method, called Markov chain model with gray relevancy weights, is applied to predict the future precipitation state by regarding the gray relevancy weights based on the special characteristics of precipitation being a dependent stochastic variable. Finally, the correction and feasibility of this model is identified by a case study.
Keywords :
Markov processes; atmospheric precipitation; weather forecasting; Markov chain model; dependent stochastic variable; gray relevancy weights; precipitation state forecasting; unascertained C-means; Economic forecasting; Engineering management; Hydrology; Meteorology; Predictive models; State estimation; Statistical analysis; Stochastic processes; Uncertainty; Weather forecasting;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Innovative Computing Information and Control, 2008. ICICIC '08. 3rd International Conference on
Conference_Location :
Dalian, Liaoning
Print_ISBN :
978-0-7695-3161-8
Electronic_ISBN :
978-0-7695-3161-8
Type :
conf
DOI :
10.1109/ICICIC.2008.426
Filename :
4603510
Link To Document :
بازگشت